The invention concerns nucleoside-5xe2x80x2-triphosphates and phosphoramidites which carry a fluorescent residue absorbing in the long wavelength range, preferably a carbocyanine group, on the base portion or on the phosphorus atom, as well as their use for labelling, detecting and sequencing nucleic acids.
Nucleic acids are of crucial importance in living nature as carriers or transferrers of genetic information. Since their discovery by F. Miescher they have therefore stimulated a broad scientific interest which has led to the elucidation of their function, structure and mechanism of action. with increasing knowledge of these fundamental molecular biological mechanisms it has in recent years become possible to pursue the new combination of genes. This technology opens for example new possibilities in medical diagnosis and therapy and in plant breeding.
An essential tool for understanding these interrelations and solving the problems was and is the detection of nucleic acids and their sequences i.e. their primary structure.
The specific detectability of nucleic acids is based on the properties of these molecules to interact, i.e. to hybridize, with other nucleic acids by forming base pairs via hydrogen bridges. Nucleic acids (probes) labelled in a suitable manner, i.e. provided with indicator groups, can thus be used to detect complimentary nucleic acids (target).
The determination of the primary structure (sequence), i.e. the sequence of the heterocyclic bases, of a nucleic acid is carried out by means of sequencing techniques. This knowledge of the sequence is in turn a prerequisite for a targetted and specific use of nucleic acids in molecular biological investigations and working techniques. The sequencing finally also utilizes the principle of specific hybridization of nucleic acids among each other. As mentioned above, labelled nucleic acid fragments are also used for this.
It is clear from the aforementioned that suitable labelling of nucleic acids is an essential prerequisite for any method of detection.
Above all, radioactive labelling with suitable isotopes such as 32p or 35S is already being used for this at an early stage. The disadvantages of using radioactive reagents are, however, obvious: such work requires special room installations and permits, as well as a controlled and complicated disposal of the radioactive waste. The reagents for radioactive labelling are expensive. A longer storage of such labelled samples is not possible due to the short half-life of the above nuclides.
Therefore in recent years there have been attempts to circumvent these serious disadvantages i.e. to get away from radioactive labelling. In doing so the high sensitivity of this type of labelling should be preserved as far as possible. Great advances have in fact been made in this case [see e.g. Nonradioactive Labeling and Detection of Biomolecules, C. Kessler (Editor), xe2x80x9cSpringer Verlag Berlin, Heidelbergxe2x80x9d 1992].
Haptens (such as biotin or digoxigenin), enzymes (such as alkaline phosphatase or peroxidase) or fluorescent dyes (such as fluorescein or rhodamine) have above all proven to be successful among others as non-radioactive indicator molecules.
Although labelling with haptens such as e.g. digoxigenin extends into the sensitivity range of radioactivity, a direct detection of hapten-labelled nucleic acids analogous to radioactive labelling is not possible. A subsequent detection reaction is necessary which is, for example, achieved by means of an antibody reaction. This indirect detection requires several steps i.e. more time and financial expense. Since proteins are used for the detection reaction, a special treatment of the solid phase (membranes, microtitre plates) by blocking and washing steps is necessary in order to reduce unspecific binding. Despite this, the sensitivity of this two-step detection is usually limited due to the occurrence of interfering background colouration resulting from unspecific protein binding. The same basically applies to direct enzyme-labelled nucleic acids.
The said disadvantage of the aforementioned indirect detection does not occur when using fluorescent-labelled nucleic acids. A direct detection is possible by exciting the fluorescence and can be visualized and measured with a suitable device (fluorescence microscope, scanner). However, the autofluorescence of cell and tissue components of the biological material to be examined such as dyes, lipids, proteins etc. also interferes in this case. Such interferences also occur particularly when using solid carrier materials (e.g. nylon membranes) due to their intrinsic fluorescence and complicate or prevent the detection.
In principle a solution to these problems is to use dyes whose excitation and emission is in wavelength ranges above 680 nm i.e. in the near infrared (NIR) range. The aforementioned interfering influences are not significant under these circumstances. A further important advantage is that very durable cheap laser diodes can be used for the excitation.
Thus for example the technique of DNA sequencing by photoelectric measurement with a laser and a sensor after fluorescent labelling of the DNA fragments is the subject matter of an application U.S. Pat. No. 4,729,947. In this method, oligonucleotides labelled with an IR dye are used in a known manner as a primer in the so-called Sanger method which act in this process as starters for the synthesis of the new complementary nucleic acid strand. However, a disadvantage of this method is thatxe2x80x94depending on the DNA to be sequencedxe2x80x94specific labelled primers have in each case to be newly synthesized again and again i.e. numerous such labelled primers. This synthesis of labelled oligomeric primers is expensive and time-consuming since the unlabelled oligonucleotide has to be synthesized at first and subsequently the signal (reporter) group is chemically attached in a second reaction.
The object of the present invention is therefore to produce compounds which enable a universal, simple and specific labelling of nucleic acids.
It is now known that nucleic acids can be newly synthesized and concomitantly labelled by the incorporation of appropriately labelled nucleoside triphosphates using polymerases. In the field of deoxyribonucleic acids (DNA) this is achieved by DNA polymerases using the methods of nick translation [Rigby, P. W. et al. (1977) J. Mol. Biol. 113, 237] and of random primed labelling [Feinberg, A. P. and Vogelstein, B. (1984) Anal. Biochem. 137, 266] by incorporating deoxynucleotides and in the case of ribonucleic acids by RNA polymerases and ribonucleotides along the lines of a transcription. A further method of labelling nucleic acids is by means of a so-called 3xe2x80x2 tailing reaction with the aid of terminal transferase and ribo or deoxyribonucleoside triphosphates.
However, nucleoside triphosphates provided with indicator molecules such as fluorescein or digoxigenin (MW 332 or 390) arexe2x80x94in contrast to their natural substratesxe2x80x94accepted relatively poorly as substrates by polymerases and incorporated relatively poorly into the newly synthesized nucleic acid (Hoeltke, H.-J. et al. (1990) Biol. Chem. Hoppe-Seyler 371, 929).
It is therefore not to be expected that indicator molecules with even considerably higher molecular weights (800-1000) would be accepted by polymerases as substrates and incorporated into nucleic acids. It is even less likely that these molecules with their given spatially demanding structure would be converted by polymerases due to strong steric hindrance.
Surprisingly it has been found that nucleoside triphosphates labelled with infrared dyes are accepted as substrates by polymerases such as T7 DNA polymerase and are incorporated into nucleic acids. The compounds according to the invention are therefore novel.
A further object of the invention is to provide a method of using the aforementioned labelled nucleotides according to the invention which enable nucleic acids labelled thus to be detected directly on solid carriers such as e.g. nylon membranes or in solution such as e.g. in microtitre plates.
As already described above, a disadvantage of labelling nucleic acids with fluorophores such as fluorescein or tetramethylrhodamine is that the intrinsic fluorescence of the carrier material interferes with the measurement of this fluorescence.
If, however, the nucleoside triphosphates labelled with IR dye according to the invention are used to label nucleic acids, these interferences are no longer significant because the measurement wavelength is located in the near infrared range.
The advantages of a nucleic acid detection by in situ hybridization are known. In this method the labelled samples or probes are either detected directly under a fluorescence microscope or, in the case of hapten labelling, are detected in an immunological reaction (ELISA) by means of a further process step. This visualization is usually achieved by immobilizing the samples for example on a nylon membrane or in a liquid homogeneous phase in microtitre plates. This additional step is very time-consuming and costly. It is thus desirable to omit this step.
In the instant invention, the possibility of directly exciting the IR-fluorescent-labelled nucleic acids by suitable laser diodes and the aforementioned insensitivity of the detection towards autofluorescence of the carrier material enables simple and cost-effective equipment to be used. The immunological detection reaction can be omitted. The IR-labelled nucleic acid is simply detected by optical means through a laser/detector combination with the aid of a suitable scanner or microtitre plate reader.
The use of the nucleoside 5xe2x80x2-triphosphates labelled with infrared fluorophores as polymerase substrates enables direct enzymatic incorporation into nucleic acids and detection of the nucleic acids labelled in this manner for sequencing, and it also allows an in situ hybridization which is also part of the invention.
The nucleoside 5xe2x80x2-triphosphates of the invention are of the general formula 
They are produced by starting with unmodified nucleosides in a well-known manner, these i.e. uridine, thymidine and cytidine in the case of pyrimidine nucleosides and the purine nucleosides adenosine and guanosine as well as the corresponding 7-deaza-purine and 7-deaza-8-aza-purine nucleosides, are chemically modified in a suitable manner at C-5 or C-6 (pyrimidine), at C-8 (purine), at C-8 (3-deaza-purine) at C-7 or C-8 (7-deaza-purine) and finally 5xe2x80x2-phosphorylated.
It is expedient that the modified group is composed of a spacer of suitable length and a terminal primary or secondary amino group which can be substituted by suitable activated fluorescent dyes (e.g. in the form of isothiocyanates or N-hydroxysuccinimide esters).
Such fluorescent dyes are used in an activated form, i.e. in a form which reacts well with for example amino groups, preferably as isothiocyanates. After the reaction is completed the fluorophores are covalently bound via NHCS groups to the modified group of the nucleotide.
The phosphorylation of the nucleosides modified in this way is carried out according to methods known in the literature [e.g. Yoshikawa, M. et al. (1967) Tetrah. Lett. 50, 5065] by reaction with phosphoryl trichloride to form the monophosphate and subsequent reaction with pyrophosphoric acid to form the desired 5xe2x80x2-triphosphate. As an alternative a direct modification of the preformed nucleoside 5xe2x80x2-triphosphates is also possible.
The fluorophores are compounds which absorb in the near infrared range i.e. between 600 and 800 nm. Those of 630 nm to 780 nm are preferred such as e.g. carbocyanines.
As already mentioned above, in addition to the method described above of incorporating labelled nucleoside triphosphates using polymerases, a further method is common which is based on the use of labelled oligonucleotides, so-called primers. As already mentioned, the conventional synthesis of these molecules in a multistep reaction is very time-consuming. In this process the signal group must be attached at the 5xe2x80x2-end of the oligomer in an additional step after the oligonucleotide synthesis is completed. Since the actual oligonucleotide synthesis is carried out in automatic synthesizers, it is extremely desirable to also be able to carry out the step of attaching the signal group already in the synthesizer. Since the oligonucleotide synthesis is composed of a stepwise attachment of monomeric building blocks, so-called nucleoside phosphoramidites, it is a further object of the invention to develop a fluorophore phosphoramidite which enables the direct incorporation of the signal group as the last step in automatic oligonucleotide synthesis. Such a NIR-dye-phosphoramidite is hitherto unknown and therefore novel.